To win at the dog track, you have to be better than the average bettor. Because pari-mutuel betting is betting against the other bettors, not the track, it's the guy or gal next to you that you have to beat, not the announcer or the track owner or managers. (Unless they're betting too, which shouldn't be allowed, but that's another article.)
Obviously, most of the crowd is wrong most of the time. I think I read somewhere once that bettors pick the winner about 1 out of 4 times. Since the favorite doesn't pay very well, that wouldn't give you a profit, that's for sure. So if you want to win money at the dog track, you're going to have to think and bet differently from how the crowd thinks and bets.
A lot of people in the crowd play numbers - the 1/2 quiniela, their kids' birthdates, their house number or their car's plate number. This is not handicapping. This is like playing the lottery, which is probably what they should do with their numbers, if they really have to bet on them.
Another segment of the crowd plays the numbers at the bottom of the program. These predictions come from the track's designated handicapper, who may or may not have a clue as to who's likely to come in. Depending on someone else to pick dogs for you, especially when the picks are shared with everyone else at the track, isn't handicapping. It's a lazy way to lose money. Same thing with getting tips from friends, trainers and anyone else who tells you about a sure thing or some set-up that they heard is happening in a race.
Einstein, who was certainly smart enough to win at the track although I don't know if he ever went, said this: Insanity is doing the same thing over and over and expecting a different result. That's what most of the crowd is doing when they try to win money at races. If you want to beat them, you have to figure out something different.
Handicap your program. Keep records. Pay attention to Class, Consistency and Competition when you pick dogs to bet. Don't look at the toteboard when you make your bets, because the toteboard only tells you what the crowd thinks the odds should be. It doesn't tell you what the odds should really be, based on sound handicapping.